Answers are grounded in this paper and cite their sources.
The use of monitoring systems for risk areas has increased significantly, especially with the use of IoT applications, which has driven Early Warning Systems (EWS). However, risk situations due to natural disasters can cause malfunction of communication networks making such monitoring unfeasible. In this paper we propose an unorthodox method for disaster detection, based on sliced network traffic monitoring. In the presented system the network traffic is monitored by a Monitoring Agent module that uses a Random Forests method to identify anomalous network traffic events. Simulation results show that our proposal is a promising solution for disaster detection, with the advantage that it could be used in different situations.
No takes yet. Share an insight, caveat, or question.
Miani et al. (2025) studied this question.